California Avocado Society 2002 Yearbook 86: 105-126
Natural Enemies Associated with Avocado Thrips in Ventura County
Avocado Groves: Results of a pilot study and year one of a three-year
survey
Pascal Oevering, Ben Faber and Phil Phillips
University of California Cooperative Extension, Ventura County
Abstract
Fourteen avocado groves were surveyed by leaf samples, beat samples and sticky
cards to assess avocado thrips (AT) and natural enemies (NE) populations in avocado
trees in 2003. In the groves, the presence of NE and AT between January and
September was only slightly affected by a single abamectin treatment, with Chrysopa
sp. being the only NE significantly impacted. In untreated groves, considerable
differences between NE populations were found. Leptothrips mali were the most
abundant predatory thrips in beat samples and arachnids the most abundant NE. On
sticky cards, Aeolothrips fasciatus were most abundant, followed by the Coleoptera,
while Ceranisus menes and Franklinothrips orizabensis were equally abundant in
untreated groves. Distance from the coast did not affect L. mali or Aeolothips, but
impacted C. menes (more abundant near the coast) and F. orizabensis (more abundant
inland). The presence of groundcover (dead, dying, mowed and living) increased the
number of parasitoids and spiders recorded in beat samples and increased the numbers
of Leptothrips mcconnelli and predaceous Coleoptera. F. orizabensis and C. menes
numbers on sticky cards declined when >70% of the orchard floor was covered. When
only living ground cover is considered, parasitoids numbers in beat samples and A.
fasciatus numbers on sticky cards increased with >70% living ground cover, while C.
menes was more abundant where living ground cover did not exceed 70% and F.
orizabensis were most abundant where 11-70% of the orchard floor was living ground
cover. The presence of flowering plants on the orchard floor had no significant effect on
NE observed in beat samples or on sticky cards. Trends indicated higher numbers of L.
mali and Orius sp. in beat samples in orchards where between 1-10% flowering plants
were present on the orchard floor, while lower numbers of F. orizabensis were found on
sticky cards in these orchards. C. menes numbers were not impacted by the presence
of flowers in the ground cover.
Most results of this survey are descriptive and need replication before conclusions or
generalizations can be made. Data from future seasons will allow for temporal analyses
which will be necessary to fully understand the NE complex in avocado orchards.
Introduction
Avocado thrips (Scirtothrips perseae Nakahara) has been a major pest in California
avocado groves since it was first found in Ventura County in 1996 (Hoddle and Morse,
1997). Most growers are currently using chemical means to control their invertebrate
pest problems and research into these areas is ongoing (Yee et al., 2001ab; Oevering
et al. 2002; Morse and Hoddle, 2003). Foreign exploration studies for natural enemies
identified six genera of predatory thrips (Aeolothrips, Aleurodothrips, Franklinothrips,
Leptothrips, Scolothrips and Karnyothrips), one parasitoid (Ceranisus) and one
predatory mite (Balaustium) associated with avocado thrips in Mexico and Guatamala
(Hoddle et al., 2002). In Ventura county, Franklinothrips orizabensis (Johansen) and
Aeolothrips kuwanaii (Moulton) were reported associated with higher numbers of
avocado thrips in a three year study from 1998-2000 (Yee et al., 2001 c). In the past
four years, possible biological control measures against avocado thrips have been
investigated, including releases of the generalist predator F. orizabensis (Oevering et
al., 2002, Hoddle et al., submitted) and releases of Chysoperla sp. (Phillips and Faber,
1999). So far none of these studies have developed into a commercially available
strategy, although some organic growers have adopted a practice of releasing
generalist predators with varying results (Oevering, personal observations).
Since reliance on biological control remains the preferred method of pest control for
most avocado growers, the focus for this study returned to natural enemies already
present in the California groves. In Ventura County, during the F. orizabensis release
studies a variety of generalist natural enemies were found on the sticky cards used to
monitor F. orizabensis in the groves (Oevering, personal observation). To identify the
spatial and temporal distribution of these generalists, a survey for natural enemies in
chemically treated and untreated groves in Ventura County was initiated.
This survey is part of a larger 3-year project that investigates the possible effects of the
presence of naturally occurring undergrowth on natural enemy abundance, where the
undergrowth may supply carbohydrates in the form of flowering plants (nectar and
pollen), shelter, and alternative food sources in the form of prey available on flowering
plants. Ultimately, this project aims to identify how cultural techniques such as pruning,
fertilization, the use of different cover crops or undergrowth mixtures can both improve
natural enemy abundance and suppress populations of avocado thrips. In this paper we
report on the results of the first year of the county-wide natural enemy survey.
Materials and Methods
Study sites and observation period
In a pilot study in 2002, avocado groves at three locations were observed. Two coastal
groves, 2.5 miles and 4.1 miles from the coast, were observed 11 times between March
and December, and 8 times between July and December, respectively. An intermediate
grove (7.3 miles from the coast) was observed 12 times between April and December.
In 2003, 14 avocado groves were selected (including those from 2002), located
between 1.6 and 24.4 miles from the coast. The groves were grouped into "coastal"
(n=3, within 5 miles from the coast), "intermediate" (n=5, between 7 and 13 miles from
the coast) and "inland" (n=6, between 16 and 25 miles from the coast). Between
January and September 2003, each grove was observed 11 to 14 times, every two
weeks between April and August (observations continued after August every 23 weeks
but are not included in this preliminary report).
Trees at all study groves were at least 90% ‘Hass’ variety, 7-15 feet tall and ranged
from 5 to 15 years of age.
Sampling for arthropods
In the pilot study, 2 sites comprised of 5 observation trees each and which were at least
10 rows apart, were observed in each grove. At each site, one sticky card (6x6 inch)
was hung underneath the canopy in each of 3 of the 5 observation trees. For each
observation, the number of immature and adult avocado thrips (AT) and any natural
enemies (ME; for a list of species observed and the abbreviated names used in paper,
see Table 1) were counted on 10 flushing leaves. The sticky cards were changed with
each observation and used cards were taken to the laboratory where the numbers of
natural enemies per card were counted.
In the 2003 survey, the only changes in observation method were additional beat
sample observations. For each quadrant in each of the 5 observation trees, a beat
sample was obtained by hitting one branch three times on a blue tray (12x16 inches).
Any natural enemies present on the tray were recorded.
Orchard ground cover
In 2003, the orchard floor was visually inspected and estimates for percentage ground
cover were recorded. Three observations were taken: (1) percentage of "all ground
cover" consisting of all organic material (living, dying, dead and mowed plants, dead
leaves and branches), (2) percentage of "living ground cover" consisting of all green
plants present including those flowering and (3) percentage of "flowering ground cover"
consisting of flowering plants only.
The subsets of observation (2) and (3) were taken for assessment of the reported
influence of alternative food and host sources on NE populations. The "all ground cover"
observation was included to verify the overall effect of cover over a bare orchard floor.
For each site, all plant species present were listed at least once during the
observations. In the future more detailed speciation of weeds will be undertaken. No
data on ground cover were collected in the pilot study.
Management information
Information on cultural, biological and/or chemical management activities were obtained
from the growers. In the pilot study in 2002, no chemical control for avocado thrips was
used in any of the 3 groves. In the 2003 survey, of the 14 groves observed, two coastal
and three inland groves were treated once with the neurotoxic insecticide/miticide
abamectin (Agri-Mek 0.15 EC, Syngenta Crop Protection, Greensboro, NC) and NR oil
for avocado thrips control.
Data presentation and statistical analyses
For the 2002 pilot study, the observations lack sufficient replication to justify statistical
analyses and the counts are presented as graphs for illustrative purposes. For the 2003
survey, statistical analyses were only performed on beat sample data and sticky trap
data for those species for which more than 16 individual insects were recorded. The
total numbers for each insect species recorded are listed in the tables for each analysis.
For lack of annual replication, temporal analyses of 2003 data have not yet been
undertaken and are not presented in this paper.
Effect of chemical treatment
All 14 groves (5 treated and 9 untreated) were used to analyze the effect of chemical
treatment using analyses of variance (ANOVA with Fisher's Least Significant Difference
multiple range tests LSD<0.05, Statgraphics© plus 5.1, 2001) on the mean number of
AT and/or NE recorded per method on both sites within each grove. Statistical trends
and significant differences are italicized in the tables. For leaf counts the mean number
recorded on 100 leaves (10 leaves on each of 10 trees) was used for analyses and
expressed in tables as mean per leaf for all observation dates pooled. Similarly, for beat
samples the mean number of NE per 40 beats (4 beats for each of 10 trees) was used
and expressed in tables as mean per grove (40 beats or 10 trees), and for sticky cards
the mean number per 6 cards was used in analyses and expressed in tables as mean
per grove.
Effect of distance from the coast
The effect of distance from the coast on numbers of AT and NE recorded was analyzed
with ANOVA (LSD<0.05) using data collected at the 9 untreated groves only (to avoid
any treatment effects of the abamectin applications). The total number of observations
per distance group is listed in each table. The mean numbers recorded from leaf counts,
beat sampling and sticky cards per observation were calculated for both sites in each of
the 9 groves as described above for the analyses of treatment effect.
Effect of ground cover
The effect of percentage orchard floor ground cover on numbers of NE recorded was
analyzed with ANOVA (with LSD<0.05) using untreated groves only. The mean
numbers recorded per site were used for those dates when estimates for ground cover
were available and beat samples and/or sticky card data were collected. Site data was
preferred over grove data, because orchard floor ground coverage at the two sites
within a grove was not always similar. The estimated percentages were grouped into
three categories for the analyses. For percentage "all ground cover" and percentage
"living ground cover" the three categories were 0-10%, 11-70% and 71-100%. For
percentage "flowering ground cover"; 0% or no flowering cover, 1-10% flowering cover
and 11-100% flowering cover were used as categories. For each of the three categories
within each of the three observations recorded on ground cover (all ground cover; living
ground cover and flowering ground cover) beat sample data were recorded per site as
number of NE per 20 beats (4 beats for each of 5 trees) averaged for all observation
dates pooled and sticky cards data were recorded as the number of NE per 3 cards
averaged for all observation dates pooled.
Results and Discussion
2002 pilot study
At both the intermediate and one of the coastal groves (coastal 1), the sticky card data
in the pilot study found the activity of C. menes out of the selection of NE observed,
most closely in synchrony with the population dynamics of AT (Figure 1). Other NE
species at these two locations were present in higher numbers only in the spring. At the
second coastal grove (coastal 2), which was used in the past for release studies of F.
orizabensis (see Oevering et al., 2002 and Hoddle et al., submitted), the numbers of C.
menes were extremely high (note Fig. 1 is scaled down and shows # per 6 cards/100 for
C. menes) in summer, although no synchrony with AT was observed. F. orizabensis
numbers increased only after C. menes numbers had dropped. This may be indicative
of an intraguild relationship, where the generalist thrips parasitoid C. menes may have
affected the population development and establishment of F. orizabensis. These
observations, amongst other considerations, prompted the 2003 survey.
Effect of chemical treatments
The use of abamectin for AT control significantly reduced the number of Anestidae, and
indicated a trend towards lower L. mali numbers in treated groves as recorded in beat
samples (Table 2). In beat samples, Chrysopa sp. showed a trend towards higher
numbers in treated groves and on sticky cards this effect was significant (Table 2).
Although all other mean numbers of NE species were higher in untreated groves, none
of these differences were statistically significant (Table 3).
When considering AT observed per leaf, fewer specimens of all stages were found in
untreated groves, although none of these differences were significant (Table 4). The
number of all NE found in leaf counts was significantly higher in treated groves (Table
4).
Abamectin is a selective "soft" chemical that has a translaminar activity within the
foliage and a short topical residual activity (Syngenta, 2001). One day after application it
has minimal effects on parasitoids (Smith et al., 1998; Brunner et al., 2001), although
the effect on predatory thrips may be more significant. F. orizabensis, and possibly
other species, occasionally supplement their diet with plant liquids and those species
may be affected by the translaminar presence of abamectin in the leaves (Hoddle,
personal communication).
A treatment suppresses AT for 60 to 90 day periods (Oevering et al. 2002). Therefore,
observations over an eight-month period (January-August) are unlikely to indicate a
treatment effect that lasted 2-3 months, especially when AT populations before
treatment often exceed those in untreated groves. When using data per observation
pooled for the total observation period, these higher pre-treatment numbers may mask
the treatment effect for a parameter that uses "mean number per observation date"
obtained from pooled data. Abamectin is still under emergency Special Local Need
(Section 18) registration for use in avocado groves and fits the needs for AT control
perfectly. Since avocado fruit is only susceptible to AT scarring for a short period in their
early development which takes place in early summer (Yee et al., 2001 d) and since AT
rarely cause economic damage to the leaves, proper timing of application will control
the AT population during the susceptible period, while allowing AT, and subsequently
NE, populations to be present at other times (Oevering et al., 2002). High temperatures
regulate AT populations in late summer, especially at inland locations (Hoddle, 2002).
Since NE associated with AT will vary with the availability of AT, the opportunistic
Chrysopa sp. may not be solely related to AT populations, considering the observations
of higher numbers in treated groves. However, small instar Chrysopa larvae have been
observed feeding on thrips including avocado thrips, and the observed higher numbers
in treated groves may indicate that the highly mobile Chrysopa adults are early arriving
NE following treatment. Temporal analyses may elucidate this enigma, but for these
analyses multiple year data are required and therefore not included in this paper.
Effect of distance from the coast
The spatial distribution of NE in Ventura County as observed using beat samples
showed that at intermediate grove distance from the coast F. orizabensis was least
abundant whilst all other species were equally abundant at all distances (Table 5). The
total numbers recorded using beat samples indicated L. mali the most abundant
predatory thrips species and the arachnids the most abundant NE (Table 5).
Using sticky cards, F. orizabensis were most often found in inland groves (Table 6). For
Chrysopa the numbers recorded in inland groves were not different from coastal groves
but higher than in intermediate groves. Onus sp. showed a trend similar to Chrysopa
sp., while C. menes was more abundant on the coast than in inland groves. Of all the
NE, the predatory thrips species A. fasciatus was most abundant on sticky cards,
followed by F. orizabensis (Table 6).
The number of immature AT per leaf in coastal groves was significantly lower than in
inland groves (Table 7). For adults a trend towards fewer numbers at intermediate
groves was found. There were no differences in the total numbers of NE in leaf counts
between coastal, intermediate and inland groves (Table 7).
These data show that the method of observation is critical when assessing populations.
Leaf observations sample only a small part of the tree. Beat samples sample NE
present on a larger amount of foliage at the moment of sampling, but (as with leaf
observations) the moment of observation may affect the NE diversity recorded, since
some NE species are active at limited times during a 24hr period. Sticky cards on the
other hand, record mostly flying NE over a longer period of time, which results in higher
numbers. Although sticky cards were positioned underneath the tree canopy to limit the
trapped species to those present within the tree, they only trap those insects that are
dispersing through the canopy, which could lead to trapping an under representation of
non-dispersing and/or aptere (wingless) species or life stages that are foraging or
stationary.
In the 2003 survey, sticky cards provided more insightful data, because statistical
differences were found that were not noticed using beat samples. However, some
species would not have been recorded without beat sample data; L. mali, the most
abundant species in beat samples, were present in very low numbers on sticky cards.
The yellow color of the sticky cards was selected for easy of handing. White sticky cards
have been reported more attractive to F. orizabensis (Hoddle et al, 2002), and although
no data is available on attractiveness for L. mali, the color may have affected the
number of L. mali on the yellow traps.
The effect of grove distance from the coast shows that there is a significant difference in
AT and NE populations between coastal and inland groves, with the intermediate
groves providing intermediate results. Coastal groves had fewer immature AT (Table 7)
and higher C. menes numbers (Table 6) whilst higher immature and adult AT numbers
at inland groves were accompanied by higher numbers of F. orizabensis. It is interesting
that higher numbers of C. menes were trapped in the 2002 pilot study (Figure 1) than in
the 2003 survey. The effect of A. fasciatus and L. mali presence in avocado groves
needs further investigation, since these species topped the list of total numbers of
predatory thrips observed by the two survey methods and neither is affected by the
grove distance from the coast. The continuation of this study for a number of years will
allow for further in-depth understanding of these insect population trends.
Effect of ground cover
Of all NE recorded in beat samples, the presence of groundcover affected only the
number of parasitoids and spiders recorded. Significantly higher numbers were found
when the orchard floor was more than 71% covered (Table 8).
On sticky cards, significantly higher numbers of Leptothrips mcconnelli (Crawford) were
found in groves with more than 71 % cover and a similar trend was observed for
predaceous Coleóptera (Table 9). F. orizabensis were found at highest numbers when
undergrowth covered between 11 and 70% of the orchard floor, while C. menes was
most abundant where no more than 10% of the orchard floor was covered (Table 9).
When only the living ground cover is considered, parasitoids were the only NE in beat
samples that were significantly affected, with increased numbers when living ground
cover exceeded 71 % (Table 10). On sticky cards, A. fasciatus were more abundant
when living ground cover exceeded 71 % of the orchard floor, while C. menes was more
abundant where living ground cover did not exceed 71 % (Table 11).
The presence of flowers, expressed as a percentage of orchard floor cover, had no
significant effect on NE observed with beat samples. Although not significant, for L. mali
and Onus sp. a trend was observed where the most insects were found in orchards
where between 1-10% flowering plants were present on the orchard floor, compared to
the other two groups of 0 % and > 11 %.
The presence of flowers, as observed using sticky cards, affected only F. orizabensis
for which a trend was found where the lowest numbers tended to be in groves with 110% flowering plants present and the highest numbers were found in groves without
flowers, although levels were not different those in groves where flowers exceeded 11
% cover of the orchard floor (Table 13).
The increased abundance of parasitoids with undergrowth is likely indicative of aphid
presence on plants in the ground cover, whether this undergrowth is living (Table 10) or
mowed (Table 8). Although parasitoids other than C. menes were not identified, aphids
were regularly seen on the undergrowth and it seems likely that the parasitoids are
associated with the presence of aphids and the honeydew they produce. Additionally,
the presence of carbohydrates and shelter may improve the habitat quality of orchards
for parasitoids. A similar situation may be occurring with the Coleoptera (Table 9). The
Coleoptera observed were for the most part Coccinellidae, which are opportunistic
predators which often prefer aphids. In the coming seasons, the presence of herbivores
on the undergrowth may need to be assessed to further develop this relationship.
Since a similar relationship between abundance of L. mcconnelli and ground cover was
found, the prey species for this completely black colored predatory thrips needs to be
identified. Foreign exploration studies found L. mcconnelli in avocado groves in Oaxaca,
Mexico, but not at other locations in Mexico, Costa Rica or Guatamala (Hoddle et al.,
2002) and information on this species is scarce.
The increased presence of spiders in trees with presence of undergrowth may also
relate to higher abundance of prey and the presence of ground cover may facilitate their
migration through the orchard habitat (Table 8). It is very likely that the parasitoids,
Coccinellidae, L. mcconnelli and spiders are not primarily associated with avocado
thrips as they are not known as exclusive thrips predators, although they may accept
avocado thrips as occasional prey. Their presence in the background NE population
may add to the effect of more specialist species.
None of the identified predatory thrips species were significantly affected by the
presence of flowers in the ground cover (Table 12 and 13), but the presence of some
living ground cover seems favorable for F. orizabensis (Table 9). Larger orchard floor
coverage with living plants increased A. fasciatus numbers (Table 11) and combined
with the trends observed for L. mali (Table 11) these results determine an interesting
dynamic where the different predatory thrips species are differentially abundant
depending upon the environment. Their presence will also vary during the year, for
which temporal analyses will be undertaken once additional seasons have been
observed.
The abundance of C. menes in the tree canopy was significantly affected by the
presence of more than 70% living ground cover in groves, where numbers on sticky
cards were lower. The group of observations in groves with more than 70% living
ground cover resulted in lower numbers of all NE, compared to the less than 70% cover,
with exception of A. fasciatus. This indicates that the presence of herbivore prey and
hosts on large (>70%) surface areas of living undergrowth may direct NE away from the
trees, while smaller (<70%) surface areas of living plants may help to build up and
sustain populations of NE in the trees. Future temporal analyses will further clarify this
relationship. The presence of Western Flower Thrips (Frankliniella occidentalis
Pergande) (Thysanoptera: Thripidae), which is common in almost all flowers in
Southern California, is not the only host responsible for the presence of C. menes
populations in the orchards, since percentage flowering plants in the undergrowth did
not affect the C. menes numbers in the trees (Table 13).
Conclusions
Most results of the 2003 survey are descriptive and need replication before any
conclusions or generalizations can to be made. The natural enemy complex present in
Ventura County avocado orchards consists of a number of species that prey on
avocado thrips. Following spiders and Anestidae in beat samples and with Coleoptera
on sticky cards, the predatory thrips species are the most abundant predators in the NE
complex. The dominant predatory thrips species varies with distance from the coast and
presence and composition of ground cover of the orchard floor. C. menes is a generalist
thrips parasitoid frequently encountered in the natural enemy complex. The abundance
of C. menes decreases with distance from the coast, and is affected by the presence of
living plants but not by the presence of flowering plants in the ground cover of the
orchard floor. The population dynamics of avocado thrips, and predators and parasitoids
in the NE complex, and relationships between these species are not analyzed in this
study. Data from future seasons will allow for temporal analyses which will be necessary
to fully understand the natural enemy complex in avocado orchards.
Acknowledgements
This survey was funded by the California Avocado Commission and the California
Avocado Society and supported by the UC Hansen Trust. We are very grateful for
access to local orchards and thank the managers/ owners Henderson, Hernandez,
Koman, McKee, Moore, Daykin, Churchill, Schaff, Steinberg, Sullivan, Syple, McGrath
and Whitney for their collaboration. A special thanks goes out to Jake Blehm, Bret
Chandler, Jane Delahoyde, Charlie Gribble, Tom Roberts and Emily Thatcher for
advice, site selection and/or data collection. This work could not have been presented
without the help of Michi Yamamoto who was instrumental in data collection and sticky
card counts.
References
Brunner, J.F., Dunley, J.E., Doerr, M.D. and Beers, E.H., 2001. Effect of pesticides on
Colpoclypeus florus (Hymentoptera: Eulophidae) and Trichogramma platneri
(Hymenoptera: Trichogrammatidae), parasitoids of leaf rollers in Washington. J.
Econ. Ent. 94(5): 1075-1084.
Hoddle, M.S. and Morse, J.G., 1997. Avocado thrips: a serious new pest of avocados in
California. Calif. Avocado Soc. Yearb. 81: 81-90.
Hoddle, M.S., 2002. Developmental and reproductive biology of Scirtothrips perseae
(Thysanoptera: Thripidae): a new avocado pest in California. Bull. Ent. Res. 92:279-
285.
Hoddle, M.S., Nakahara, S. and Phillips, P.A., 2002. Foreign Exploration for Scirtothrips
perseae (Thysanoptera: Thripidae) and associated natural enemies on avocado
(Persea americana Miller). Biological Control, 24: 251-265.
Hoddle, M.S., Oevering, P., Phillips, PA. and Faber, B.A. Evaluation of augmentative
releases of Franklinothrips orizabensis for control of Scirtothrips perseae in
California avocado orchards. Biological Control, submitted
Hoddle, M.S., Robinson, L. and Morgan, D. 2002. Attraction of thrips (Thysanoptera:
Thripidae and Aeolothripidae) to colored sticky cards in a California avocado
orchard. Crop Protection 21 (5): 383-388.
Morse J.G., and Hoddle, M.S., 2003. Avocado thrips biology and control. AvoResearch,
April 2003, 1-8.
Oevering, P., Hoddle, M., Faber, B.A., and Phillips, PA. 2002. Release of
Franklinothrips orizabensis for control of Scirtothrips persea, Proceedings of the 3rd
California Conference on Biological Control, pesticide efficacy trials, AvoResearch,
April 2002, 1-8.
Phillips, PA. and Faber, B.A., 1999. Avocado thrips subproject 2: pesticide evaluation
and phenology in the field. Proc. Calif. Avocado Research Symposium: 27-29.
Smith, A., Smith, N.J. and Smith K.M., 1998. Effect of abamectin on citrus rust mite
Phyllocoptruta oleivora and brown citrus rust mite Tegolopgus australis and the
scale natural enemies Aphytis lingnanensis and Chilocorus circumdatus on
oranges. Plant Protection Quarterly 13(3): 136-139.
Syngenta, 2001. Agri-Mek miticide/insecticide technical bulletin. SCP 801 00053-C
3/01.
Yee, W., Phillips, PA. and Faber, B.A., 2001 a. Effects of aerial spray volume, coverage
and Sabadila on Scirtothrips perseae (Thysanoptera: Thripidae). J. Econ. Ent. 94:
1085-1089.
Yee, W., Phillips, P.A. and Faber, B.A., 2001 b. Effects of aerial spray volume, coverage
and Abamectin on Scirtothrips perseae (Thysanoptera: Thripidae). J. Econ. Ent. 94:
1090-1096.
Yee, W., Phillips, PA., Rodgers, J.L. and Faber, B.A., 2001 c. Phenology of arthropod
pests and associated natural predators on avocado leaves, fruit and in leaf litter in
Southern California. Env. Ent. 30 (5): 892-898.
Yee, W., Phillips, P.A., Faber, B.A., and Rodgers, J.L. 2001 d. Relationships between
Scirtothrips persea (Thysanoptera: Thripidae) populations on avocado leaves, fruit
and scarring damage on fruit. Env. Ent. 30(5): 932-938.